Minutes Ago Privacy Policy

Last updated: 2026-11-03
Provider: Kurokoto Lab (Yuji Kuroko)
Contact: [email protected]
Published at: https://www.kurokoto.com/legal/minutesago/privacy

This English translation is provided for reference only. The Japanese version is the official text and prevails in case of any discrepancy.

1. Summary

Minutes Ago (the "App") records, transcribes, identifies speakers and writes meeting minutes on your Mac. Kurokoto Lab (Yuji Kuroko; "we" / "us") does not collect your personal data through the App. The App contains no analytics, advertising or tracking, and has no accounts. Purchases of the paid feature "Minutes Ago Pro" are handled by Apple in the App Store version and by Polar in the direct version (sections 3 and 4).

2. Data the App handles on your Mac

Data	Where it is stored	Retention and deletion
Microphone audio and the audio of the app you choose (app audio)	In the App's memory (default). When the App quits, it writes the recent audio inside the App's container and picks it up at the next launch. Inside the App's container (`~/Library/Containers/<the App>/`) when you choose the SSD in the settings, and during a meeting	Deleted automatically after the retention period you set (30 minutes to 24 hours, 90 minutes by default). During a meeting, everything from its start is kept until the meeting ends (up to 24 hours). "Erase Live Audio…" in the menu deletes it immediately
Transcripts, speaker labels and the names you give, and markers (their times and the names you give them)	Same place (an SQLite database)	Deleted with the audio, on the same schedule. "Erase Live Audio…" deletes them immediately
Files you export (audio, transcripts, minutes)	The folder you choose	Never deleted by the App; you manage them
Saved recordings (what you save when a meeting ends or from a range in the Recordings & Minutes window, and what you create with audio file import (the audio file import row below): audio, transcript (and, if you transcribe it with other models, one transcript per model), speaker labels and names, markers, screenshots, minutes, and the name you give it)	Inside the App's container (the `library/` folder, one folder per recording)	Not deleted by the retention period; kept until you delete them. Delete them one by one from the list of recordings in the Recordings & Minutes window, or all at once with "Delete All Saved Recordings…" in the settings. "Erase Live Audio…" does not delete them
Screenshots (images of the window of one app you choose, taken during a meeting at the interval you choose, 3 to 60 seconds, and when you choose Add Screenshot; JPEG)	During a meeting, inside the App's container (the `shots/` folder). When the meeting is saved, they move into that recording's folder	Deleted when you discard the meeting, or choose "Erase Live Audio…" during the meeting. Once saved, kept until you delete the recording
The audio of the audio and video files you choose for audio file import (Minutes Ago Pro), and their transcripts, speaker labels and names, and minutes	The file's audio, converted (AAC; a video's picture is not kept), and its transcript are kept as a saved recording inside the App's container (the `library/` folder), together with its speaker labels and names and minutes. The recording is named after the original file (without its extension), which you can rename, and dated with the file's date (such as its creation date). The original file stays where it is (the App only reads it and does not change it; it keeps neither the original file itself nor its location)	The same as the Saved recordings row above (kept until you delete them). If you cancel an import or it fails, that recording is deleted. If an import ends partway, for example because the App was forced to quit, an audio-only recording may remain (you can delete it like any saved recording). You manage the original file
Temporary audio files while transcribing again (transcribing a selected range again, and transcribing with Parakeet TDT ja or Cohere Transcribe)	Inside the App's container (the `tmp/` folder)	Kept only during that transcription and deleted when it ends (also when it is cancelled or fails). If the App quits during it, they are deleted at the next launch. Transcribing a selected range again is also stopped by "Erase Live Audio…", which deletes its temporary file
Settings (including minutes presets and the names, model names and server URLs of AI providers) and the replacement dictionary	The App's preferences (UserDefaults, in the container)	Same as the container (see section 7)
The API keys of external AI providers (item 4 of section 3), one per provider, and the agent access token	This Mac's Keychain	Can be removed in the App's settings (removing a provider also removes its API key)
The number of free tries used, the last confirmed purchase state (App Store version), the license key and activation details (direct version)	This Mac's Keychain	For free tries, only the number used and the dates are kept, not which operation or range they were used for. They remain after the App is deleted, so that free tries are not reset. The direct version's license details are removed by "Deactivate This Mac"
Update settings (direct version only: whether to check automatically, the time of the last check)	The App's preferences (UserDefaults, in the container)	Same as the container (see section 7)
The connection point for the command line tool (direct version only; a socket)	Inside the App Group container (`~/Library/Group Containers/UY6Q9BGBWY.com.kurokoto.minutesago/`)	Created only while the App is running and removed when it quits. No data is stored there (an empty folder may remain after the App is deleted)
The connection point for agent access (App Store version only; a socket)	Inside the App's container (`~/Library/Containers/com.kurokoto.minutesago/Data/`)	Created only while agent access is on, and removed when you turn it off or the App quits. No data is stored there

Recording starts only after you choose Start Recording in the onboarding, and you can pause or erase at any time from the menu. While recording, the menu bar icon and macOS's microphone indicator show it.
Transcription uses Apple's speech recognition (SpeechAnalyzer) on device, speaker identification uses Nemotron 3 Diarization (Core ML) on device, and minutes are written by Apple Intelligence (on-device Foundation Models) by default. When you choose it, transcribing again (a selected range or a saved recording) and audio file import also use Whisper, Parakeet TDT ja or Cohere Transcribe (Core ML) on device.
Minutes are shown on screen and become a file only when you choose to save them. Minutes of a saved recording are also kept in that recording's folder.
Screenshots are taken only when you turn on "Take screenshots of" and choose an app and an interval when starting a meeting (off by default). They need macOS's Screen Recording permission, and the App asks for it only when you start a meeting from the menu. After you allow it, macOS may still ask from time to time whether the App may keep recording the screen. While screenshots are being taken, the menu bar icon shows a red dot and the menu shows how many have been taken. Only the chosen app's window is captured, never the whole screen or other apps. During a meeting with screenshots, one more is also taken at the moment you choose Add Screenshot (from the menu, the meeting bar, a key you set, or an agent you connected). A screenshot identical to the last one kept is not saved ("Skip identical screenshots" in the settings, on by default). During a meeting, its name, start time and screenshot settings are kept in the App's container (meeting.json) so that it can resume after the App restarts (deleted when the meeting ends).
If you turn on "Include screenshots" in the minutes settings (off by default), screenshots are also used to write the minutes. With on-device minutes (Apple Intelligence), the screenshots are read on this Mac and nothing leaves it. With a minutes engine outside the Mac, see items 3 and 4 of section 3.
Audio file import also transcribes, identifies speakers and writes minutes on this Mac in the same way (unless you choose item 3 or 4 of section 3 as the minutes engine). The imported file's audio (converted) and its transcript are kept inside the App's container as a saved recording (the Saved recordings row of the table above). The original file is not changed. Nothing imported is written where the live audio and transcripts are stored (the first two rows of the table above).

3. Connections off your Mac

The App connects to anything outside your Mac only in these 9 cases (1 to 7 apply to the App Store version; 1 to 6, 8 and 9 apply to the direct version):

Speech model downloads: macOS downloads language models from Apple (Apple's privacy policy applies). No audio or transcripts are sent.
Speaker and transcription model downloads: the first time you use Identify speakers, or when you choose to download it in the Models settings, the App downloads the speaker model (about 200 MB) from Hugging Face (huggingface.co and the hosts it serves files from). Transcription models (Whisper, about 1.6 GB or about 650 MB; Parakeet TDT ja, about 620 MB; Cohere Transcribe, about 2.2 GB) and the speech detection model downloaded together with Cohere Transcribe (Silero VAD, about 1 MB) are downloaded from Hugging Face (Whisper from argmaxinc/whisperkit-coreml, and its tokenizer from openai/whisper-large-v3; Parakeet TDT ja from FluidInference/parakeet-0.6b-ja-coreml; Cohere Transcribe from FluidInference/cohere-transcribe-03-2026-coreml; the speech detection model from FluidInference/silero-vad-coreml) over HTTPS only when you choose to download them in the Models settings. No audio or transcripts are sent in either case, but like any download, your IP address and similar connection data reach that server.
If you choose Apple Private Cloud Compute for the minutes: the transcript text of the selected range, of the file in audio file import, or of the saved recording (never audio), and, if you turn on Include screenshots, the screenshots of that range, are sent to Apple's Private Cloud Compute to write the minutes. The App asks you before sending. Apple's privacy policy applies.
If you choose an external AI provider for the minutes: only when you set up an external AI provider (OpenAI, Anthropic, Google (Gemini), or an OpenAI-compatible server you specify) and generate minutes with it, the transcript text of the selected range, of the file in audio file import, or of the saved recording (never audio; and, if you turn on Include screenshots, the screenshots of that range) is sent to that provider. The App asks you before sending and names the destination (unless you chose "Don't show again" for that destination; you can turn the question back on in settings. Sending with screenshots is confirmed separately from sending the transcript alone). When you choose Test Connection in settings, the App sends one short request without any transcript, after the same question. Your own API key is used (it is kept in this Mac's Keychain and never sent to us). How the provider handles the data is governed by your agreement with it and its privacy policy; we are not involved.
Agent access (off by default): only when you turn it on, the App gives transcripts, audio and saved recordings (including screenshots) to apps on this Mac (for example AI agents). In the direct version, an MCP server listens on 127.0.0.1 on this Mac and serves apps that hold the token. The App Store version does not listen on the network: the App's own executable, which you register with an agent and which the agent starts as its MCP server, passes them on through a connection point inside the App's container (see section 2). That connection point accepts requests only from programs run by the same user on this Mac that hold the token (the App itself). The App itself sends nothing off the Mac, but an agent you connect may send that data to outside services. You decide which agents to connect. In the direct version, the connection point for the included command line tool (minutesago; see section 2) also accepts requests from programs run by the same user on this Mac whenever the App is running. It does not communicate outside the Mac either.
Shortcuts (App Intents): transcripts, audio and minutes (including the transcripts of saved recordings) are passed to the shortcuts you create and run. What happens next depends on those shortcuts.
App Store purchases (App Store version): when you buy Minutes Ago Pro, restore purchases and the like, macOS's App Store system communicates with Apple. No audio or transcripts are sent. Information about the purchase is handled under Apple's privacy policy.
License activation (direct version only): when a license is activated, checked periodically or deactivated, the App sends the license key, our organization and product identifiers, the activation identifier and this Mac's name (when activating, so that you can tell your Macs apart in Polar's customer portal) to Polar (api.polar.sh). No audio or transcripts are sent. Like any connection, your IP address and similar connection data reach that server. The App Store version never makes this connection.
Update checks (direct version only): when you check for updates, or about every 24 hours if you allow automatic checks, the App asks GitHub (github.com, objects.githubusercontent.com) whether a new version is available. The request includes the App's name and version and the version of its update feature (Sparkle), and like any connection, your IP address and similar connection data reach GitHub. New versions are also downloaded from GitHub. No audio, transcripts, usage data or information about your Mac's configuration is sent. You can decline automatic checks when first asked and turn them off at any time. The App Store version never makes this connection.

4. What we receive

We receive no audio, transcripts, minutes, settings or usage data. The number of free tries used is not sent off the Mac either.
Crash reports and diagnostics may reach us through Apple only if you agreed to "Share with app developers" in macOS. Their contents and sharing follow Apple's system and your settings.
Apple handles App Store purchases; we receive no personally identifying information about you.
Polar handles purchases of the direct version. In Polar's dashboard we can see orders, license keys and the list of activations (including the Mac names sent under item 8 of section 3), and we use them to manage licenses, process refunds and answer inquiries.

5. Sharing with third parties

We share none of your data with third parties. Data goes to the destinations in section 3 (Apple, Hugging Face, an AI provider you choose (OpenAI, Anthropic, Google, or a server you specify), agents you connect, and Polar and GitHub in the direct version) only as a result of your actions.

6. People who are recorded

The App records people around you and in your meetings. Some countries and regions require the consent of everyone in a conversation before it is recorded. Tell the other participants that you are recording, and follow the laws and rules that apply to you (the App shows this reminder in the onboarding and before the first export). If you take screenshots during a meeting, they can show participants' faces and names, chat, shared documents and notifications. Tell the other participants that the screen is being captured too.

7. Your choices and deletion

You can pause and resume recording, change the retention period and where live audio is kept (memory or SSD), and turn app audio, transcription, the minutes engine and agent access on or off in the menu and settings.
You can withdraw microphone and system audio recording permission at any time in System Settings › Privacy & Security. You can withdraw the Screen Recording permission (screenshots) in the same place, under Screen & System Audio Recording.
You can delete saved recordings one by one from the list of recordings in the Recordings & Minutes window, or all at once with "Delete All Saved Recordings…" in the settings. Deleting a recording deletes its whole folder, including the transcripts from every model. Deleting lines from a saved recording's transcript removes them only from the transcript of the model shown at the time; that recording's transcripts from other models and its audio remain.
You can delete downloaded models (transcription and speaker models) in the Models settings.
"Erase Live Audio…" deletes the buffered audio (both in memory and in the container), transcripts and markers immediately (during a meeting, it also discards the meeting and its screenshots). "Erase Live Audio…" does not delete saved recordings. On macOS, moving the App to the Trash may leave its container (~/Library/Containers/<the App>/) behind, so run "Erase Live Audio…" and "Delete All Saved Recordings…" before deleting the App, or delete the container too (exported files stay in your folder).

8. Children

The App is not directed to children.

9. Changes

When this policy changes, the date above is updated and the new version is published on our website (https://www.kurokoto.com/legal/minutesago/privacy).

10. Contact

Questions about this policy: [email protected]

Last updated: November 3, 2026
Kurokoto Lab